516 NOTES

containing an excess of iron (Mueller and Miller, J. Immunol., in press). For

purposes of attempted purification it has been grown in a heart infusion medium
containing glucose, tryptic digest of casein, and an excess of reduced iron. The
crude toxins obtained from this medium contain 200,000 to 800,000 mouse
M.L.D. per ml (ca. 5 X 104 mg N2 per M.L.D.). Several lots of the crude
toxin have been purified by precipitation with cadmium chloride according to
Eaton and Gronau, by precipitation with 0.5 saturated ammonium sulfate
followed by dialysis at 0 to 5 C against 0.9 per cent NaCl, or by precipitation
with cadmium chloride and reprecipitation with ammonium sulfate. The best
preparation (table 1, toxin no. 4) contained 0.023 X 10- mg N2 per M.L.D.
The relatively constant ratio, Lf/M.L.D., indicated that losses during purification
did not result from inactivation of toxin; rather, losses resulted largely from
incomplete elution from the cadmium precipitate.
Although the N2/M.L.D. ratio represents a toxin of a hundredfold greater
purity than any previously obtained, we have as yet no evidence that a limit of
purification has been reached. Further, these data do not yet exclude the pos-
sibility that the toxic entity is nonprotein and that it is carried through the
purification procedures by adsorption on inactive protein. Indirect evidence,
however, argues against this assumption: 80 per cent of the total nitrogen of the
purified preparations is precipitated by 5 per cent trichloroacetic acid and 70
per cent by 1 per cent metaphosphoric acid; and the total nitrogen, calculated
as protein, constitutes .100±2 per cent of the dry weight in the purified
preparations.

TRIPLE-SUGAR IRON AGAR MEDIUM FOR THE IDENTIFICATION

OF THE INTESTINAL GROUP OF BACTERIA
A. A. HAJNA
Bureau of Bacteriology, Maryland State Department of Health, Baltimore, Maryland
Received for publication March 2, 1945
Kligler's iron agar medium has replaced Russell's and Krumwiede's medium
in many laboratories as a primary differential medium for enteric pathogens.
The advantage of Kligler's medium over Russell's and Krumwiede's lies in the
reaction for H2S. Since practically all members of the Eberthella and Salmonella
genera produce H2S whereas those of the genus Shigella do not, this reaction is
of great practical importance in combination with the characteristic differential
reactions of acid and gas production first suggested by Russell, and by means
of which enteric pathogens can be distinguished from coliform and other non-
pathogenic bacteria. The value of Kligler's medium has been limited by the
fact that H2S is not produced by all strains of Eberthella typhosa and by certain
members of the genus Salmonella. Production of H2S is sometimes so slow
that reactions may be negative at the important 18 to 24-hour incubation period
and the reactions are not always so clear-cut as is desirable.
A new triple-sugar iron agar (TSI) medium has been developed which gives
 NOTES 517
more satisfactory reactions, that is, reactions which are more clear-cut for acid
and gas, and more sensitive for H2S. This is important since selection of car-
bohydrate test media to be used for preliminary identification of members of
the Eberthella, Salmonella, or Shigella genera should be based on the TSI reac-
tions. The formula for the TSI medium is given below.
To each liter of distilled water, add
Agar (dry) ................................................... 13.0 g
B.B.L. nutripeptone* ................................................... 20.0 g
Sodium chloride ......................................................... 5.0 g
Dissolve in the Arnold sterilizer or in flowing steam in the autoclave. When melted-,
adjust to pH 7.5.
Admix
Lactose . .......................................................... 10.0 g
Sucrose ......................................................... 10.0 g
Glucose . .......................................................... 1.0 g
Sodium thiosulfate ................................................. 0.2 g
Ferrous ammonium sulfate ............................................... 0.2 g
Check the pH, which should then be 7.4.
Lastly, add
Phenol red (1% aqueous solution) ........................................ 2.5 ml.
Dispense approximately 5 ml. in 15 x 100 mm tubes. Autoclave at 12 pounds pressure
for 15 to 17 minutes.
* Bacto-peptone 15 g plus bacto-proteose peptone 5 g gives equally good results.

Sodium sulfite and meat extract have not been found to increase materially
the production of H2S. The role of sucrose in the medium is to eliminate cer-
tain sucrose-fermenting, non-lactose-fermenting bacteria of the Proteus and
paracolon groups.
All of the Salmonella cultures listed in Circular 54 of the University of Ken-
tucky, plus 26 types recently furnished by Dr. P. R. Edwards, were used in
this investigation. Very few Salmonella organisms failed to produce H2S,
notabily S. paratyphi A, S. abortus-equi, S. typhi-suis, S. berta, S. sendai, and
S. papuana. S. cholerae-suis, hitherto distinguishable from S. cholerae-suis var.
kunzendorf by its failure to produce H2S, produced it in the medium described
here, though in lesser quantity than did the Kunzendorf variety. Eberthella
typhosa and S. gallinarum are indistinguishable in this medium.